Body dysmorphic disorder (BDD), a preoccupation with an imagined or slight defect in appearance, is a chronic, disabling, and costly disorder that is probably far more common than is recognized. This often-secret disorder causes severe distress and impaired functioning, and can lead to repeated hospitalization and suicide. Despite its recognition for more than a century, BDD has received little empirical study. In particular, there have been no controlled treatment trials of this disorder. Although neuroleptics are often used to treat BDD, recent open treatment trials suggest that BDD may respond preferentially to serotonin-reuptake inhibitors (SRIs). These preliminary studies also suggest that delusional BDD responds as well to SRIs as does nondelusional BDD. However, these findings require confirmation in a placebo-controlled trial. The proposed study compares the efficacy of fluoxetine versus placebo in a 12-week randomized double-blind trial in 96 outpatients with BDD or its delusional variant (delusional disorder, somatic type). It will assess in an exploratory manner whether treatment response is predicted by the presence and degree of delusional thinking; other predictors of treatment response will also be assessed. The study will also investigate whether BDD and its delusional disorder variant are separate disorders or the same disorder by comparing demographic and phenomenologic features, onset and course, associated psychopathology, family history, and treatment response. The proposed study will provide needed data on the treatment of a chronic, debilitating, and costly, yet understudied, disorder. It will also provide data on the question of whether BDD and its delusional variant are different disorders or the same disorder--a controversial classification issue that was debated for DSM-IV and remains to be resolved for DSM-V.